The visualization here is excellent, and I love watching how a link spreads through the little ecosystem of linkers of different sizes.
However, I am very concerned about the headline, which suggests that HN is the most interesting part of this. I fear that posts framed this way are the first step into the pit of self-referentiality that reddit is permanently stuck in.
Did you notice that the user zachbeane that submitted the story is the same Zach Beane that wrote the story and the visualization library? Maybe he wrote the title that way to make it more relevant to HN readers.
Fantastic visualization of that data. You should release some sort of library for creating these from Analytics data or however you were logging the referrals.
I would love to have a page with my traffic displayed this way - maybe over a few different time scales.
It is always nice to see such well thought out views of data like this, for movies and web traffic it really seems to do the task.
Title is inaccurate. While the chart shows it, the linked article is not about reddit vs. hacker news traffic, nor does it even mention "Hacker News." It is a visualization of traffic from several sources, one of which is HN.
What's so clever about his implementation is that he took the stream graphs and made the rendering very simple (blocks in columns linked by bezier curves, most drawing kits make these trivial). It turns out that it's actually an improvement for lots of different datasets.
It's really a great hack, and I suspect that a lot of people will copy it shortly. Certainly, I already am.
I really like the design sense here. But the decision to give some whitespace padding between items (sites, movies, etc.) means that you can't get an accurate idea of the totals for a given period. Compare to these: http://www.leebyron.com/else/streamgraph/
For the movie graphs, the visualizations are limited to the top twenty-five movies so it doesn't matter. But his traffic visualization shows all referrers. Those graphs would convey more information without the whitespace.
The most interesting line to follow was Google. It was never at the top, but it was a snake slithering through all the results, patiently contributing at least 50% of the traffic.
No offence, but the numbers are simply too low to read anything meaningful into them. I'm guessing you never even reached page three of Reddit, with less than 700 hits.
I don't see how you can possibly draw any conclusions from this. For example, if you'd reached close to the top on Reddit, would the figures progress in the same way? I very much doubt it.
The "hits" number is the number of hits in that hour. Move your mouse left and right, and you'll see he got 376 in the first hour, 565 in the second, 721 in the next, and so on.
However, I am very concerned about the headline, which suggests that HN is the most interesting part of this. I fear that posts framed this way are the first step into the pit of self-referentiality that reddit is permanently stuck in.